Publication number  US7397916 B2 
Publication type  Grant 
Application number  US 10/433,966 
PCT number  PCT/CA2001/001729 
Publication date  Jul 8, 2008 
Filing date  Dec 10, 2001 
Priority date  Dec 8, 2000 
Fee status  Lapsed 
Also published as  CA2327911A1, CA2431443A1, EP1350154A2, US7809135, US20040139340, US20060140401, WO2002046890A2, WO2002046890A8 
Publication number  10433966, 433966, PCT/2001/1729, PCT/CA/1/001729, PCT/CA/1/01729, PCT/CA/2001/001729, PCT/CA/2001/01729, PCT/CA1/001729, PCT/CA1/01729, PCT/CA1001729, PCT/CA101729, PCT/CA2001/001729, PCT/CA2001/01729, PCT/CA2001001729, PCT/CA200101729, US 7397916 B2, US 7397916B2, USB27397916, US7397916 B2, US7397916B2 
Inventors  Harold J. Johnson, Stanley T. Chow, Philip A. Eisen 
Original Assignee  Cloakware Corporation 
Export Citation  BiBTeX, EndNote, RefMan 
Patent Citations (6), NonPatent Citations (6), Referenced by (44), Classifications (20), Legal Events (4)  
External Links: USPTO, USPTO Assignment, Espacenet  
This application claims the priority of Canadian Patent Application Serial No. 2,327,911, filed Dec. 8, 2000; and PCT Application Ser. No. PCT/CA01/01729, filed Dec. 10, 2001, which are incorporated herein by reference.
The present invention relates generally to computer software and electronic hardware, and more specifically, to a method, apparatus and system resistant to a “white box attack”; that is, a system which will protect certain information from discovery even when the attacker has total visibility into software implementation and execution.
The use of computers and computer software in all of their various forms is recognized to be very common and is growing everyday. In industrialized nations, hardly a business exists that does not rely on computers and software either directly or indirectly, in their daily operations. As well, with the expansion of powerful communication networks such as the Internet, the ease with which computer software programs and data files may be accessed, exchanged, copied and distributed is also growing daily.
In order to take advantage of these computer and communication systems and the efficiencies that they offer, there is a need for methods of storing and exchanging computer software and information securely. Information security issues can generally be categorized as one of the following:
One method of maintaining confidentiality or privacy that has demonstrated widespread use and acceptance is encryption of data using secret cryptographic keys. Such methods are generally accepted as secure, as an attacker must perform an impractically large number of mathematical tests to identify the cryptographic key required to decode a given encrypted data file. Cracking the Data Encryption Standard (DES) for example, would require an average of 2^{55}different keys to be tested, requiring more than 1 thousand years of testing at a rate of one million key tests per second. DES is just one of several block cipher methods which are very fast and are widely used—block ciphers are schemes in which data is divided up into blocks which are encrypted and decrypted separately from one another. If the cryptographic key is kept secure, it offers very good security.
There are many applications, however, in which the encryption key or other secret information (such as biometric data) must be hidden in application software itself. One such example is a Smart Card. Smart Cards are credit cardsized devices which have a small amount of electronic memory and a small microprocessor. They are often used in electronic commerce applications or to record personal information such as health records.
Existing encryption systems are designed to protect their secret keys or other secret data against a “black box attack”. This is a situation where the attacker has knowledge of the algorithm and may examine various inputs to and outputs from the algorithm, but has no visibility into the execution of the algorithm itself. Typical black box attacks are categorized as follows:
While such attacks are powerful enough by themselves, it has been shown that the black box model does not reflect reality. Often, if the attacker has sufficient access to the targeted algorithm to mount an adaptive chosenplaintext attack, he is also in a position to observe at least some aspect of the execution of the algorithm itself.
For example, in their paper on Differential Power Analysis (DPA) titled: Introduction to differential power analysis and related attacks, 1998 (available on the Internet at http://www.cryptography.com/dpa/technical), Paul Kocher, Joshua Jaffe and Benjamin Jun demonstrated that Smart Cards protected with DES could be cracked in a matter of minutes rather than the theoretical thousand years of testing. DPA merely monitors the power consumed by the Smart Card while it is executing the DES algorithm on a chosen input. The powerconsumed by the Smart Card at any given point in time corresponds with the bias state of the transistors in the device. The DPA attack can thus deduce the number of bits that are correct in a proposed key by statistical analysis, converging to the correct key very quickly.
The DPA attack shows that having very limited access to the execution of an algorithm designed to defend against a black box attack, is sufficient to make that algorithm completely insecure. Therefore, encryption algorithms must be designed to be secure against a much more powerful attack model—the “white box attack”. A white box attack is simply an attack on a software algorithm in which the attacker has full visibility into the execution of the algorithm (note that the DPA attack may be characterised as a “grey box attack” because the attacker is only able to observe a small part of the execution).
Unfortunately, existing encryption and decryption algorithms and related algorithms including those for digital signatures, password authentication, Smart Cards and the like, are not resistant to white box attacks. In fact, many believe it is impossible to achieve such protection.
Several approaches have been suggested, but offer very weak protection. For example:
As well, in their patent application serial no. PCT/US98/12017, Christian Collberg, Clark Thomborsson and Douglas Crow present slightly more sophisticated variations on these themes. While these techniques might sufficiently obscure a key (for example) against a generic, automated attack, they are not resistant to a rigorous white box attack. In most cases, simply tracing execution of the software code will reveal the encryption key and any secure data completely.
Collberg et al, for example, suggest making some simple linear transformations to data values. A careful tracing of the program execution will identify how these transformations are done and undone. As well, since these transformations are linear, they can be reduced.
In addition, a variety of cryptographically weak approaches have been used for encryption and decryption, which avoid the use of any explicit key whatever. These methods are vulnerable either to a cryptographic blackbox attack if plaintext can be recognized in an automated way, or to algorithmic analysis with the aid of debugging tools, since the wouldbe encryption is then a data transformation of quite limited algorithmic complexity.
In general, then, the state of the art has been that programs could not be made effectively secrethiding and tamperresistant. In particular, cryptographic keys for reasonably secure ciphers could not be securely hidden in software.
There is therefore a need for a method, apparatus and system for encryption that is tamperresistant, allowing secret cryptographic keys, biometric data and encrypted data to have software operations executed upon it, without fear that security will be breached.
It is therefore an object of the invention to provide a method and system that improves upon the problems described above.
One aspect of the invention is broadly defined as a method of modifying software algorithms to foil tracing and other static, dynamic, and statistical attacks comprising the steps of: encoding the software algorithm; and widely diffusing sites of information transfer and/or combination and/or loss.
Another aspect of the invention is defined as a method of protecting computer software comprising the steps of: identifying functions and transforms substantive to the targeted software program; generating new functions and transforms which alter the processing activity visible to the attacker; and replacing those identified functions and transforms with the new functions and transforms in the software program.
Another aspect of the invention is defined as an apparatus for modifying software algorithms to foil tracing and other static, dynamic, and statistical attacks comprising means for encoding the software algorithm; and means for widely diffusing sites of information transfer and/or combination and/or loss.
A further aspect of the invention is defined as a computer readable memory medium, storing computer software code executable to perform the steps of: encoding the software algorithm; and widely diffusing sites of information transfer and/or combination and/or loss.
An additional aspect of the invention is defined as a computer data signal embodied in a carrier wave, the computer data signal comprising a set of machine executable code being executable by a computer to perform the steps of: encoding a software algorithm; and widely diffusing sites of information transfer and/or combination and/or loss.
These and other features of the invention will become more apparent from the following description in which reference is made to the appended drawings in which:
As noted above, most security software is designed under the assumption that the software will be applied in a secure environment, that is, in a blackbox model. This of course, is not realistic, and as a result, most security software cannot withstand a concerted attack.
The central failing of blackbox security models is that the specific implementation of the software algorithm is considered to be irrelevant to security. This is in direct contrast to the method of the invention. The invention appreciates that in the case of a whitebox model, the algorithm is critical, and changing the specific implementation of the algorithm is the primary means for providing security.
The invention provides ways to make finding an embedded cryptographic key or other hidden information combinatorially difficult for the attacker, even under this severe threat model. Such methods are inherently bulkier and slower than blackbox cryptography, but there are many applications in which the tradeoff is well worthwhile, including, for example, Smart Card and their hardware alternatives.
In broad terms, the method of the invention which addresses the objects outlined above, is presented as a flow chart in
A number of different techniques for effecting the invention are described hereinafter. These techniques may be grouped generally as follows:
Hence, the invention can be employed to protect any manner of software from being analysed, reversedengineered, or simply observed to discover secure data such as cryptographic keys. Cryptographic keys can then be incorporated into software programs without the danger of the cryptographic key being disclosed, or the program being altered to do anything other than what it was originally intended to do. Executable music files, for example, can be bound to a particular processing device or to a password, and attackers are unable to modify the code to allow it to be used by others or on other devices.
Similarly, passwords, biometric data and other secure programs and data files can also be securely stored, transferred and executed using the method of the invention.
Being a software solution, the cost of the invention is very small and the invention can be transported electronically. The invention has none of the costly administrative and physical limitations of hardware solutions. The degree of complexity of the invention is easily scalable, so that the degree of analysis required to overcome it can be made impractically great.
An example of a system upon which the invention may be performed is presented as a block diagram in
The computer 20 may contain one or more processors or microprocessors, such as a central processing unit (CPU) 24. The CPU 24 performs arithmetic calculations and control functions to execute software stored in an internal memory 26, preferably random access memory (RAM) and/or read only memory (ROM), and possibly additional memory 28. The additional memory 28 may include, for example, mass memory storage, hard disk drives, floppy disk drives, magnetic tape drives, compact disk drives, program cartridges and cartridge interfaces such as those found in video game devices, removable memory chips such as EPROM or PROM, or similar storage media as known in the art. This additional memory 28 may be physically internal to the computer 20, or external as shown in
The computer system 14 may also include other similar means for allowing computer programs or other instructions to be loaded. Such means can include, for example, a communications interface 30 which allows software and data to be transferred between the computer system 14 and external systems. Examples of communications interface 30 can include a modem, a network interface such as an Ethernet card, a serial or parallel communications port. Software and data transferred via communications interface 30 are in the form of signals which can be electronic, electromagnetic, optical or other signals capable of being received by communications interface 30.
Input and output to and from the computer 20 is administered by the input/output (I/O) interface 32. This I/O interface 32 administers control of the display 16, keyboard 18, external devices 22 and other such components of the computer system 14.
The invention is described in these terms for convenience purposes only. It would be clear to one skilled in the art that the invention may be applied to other computer or control systems 14. Such systems would include all manner of appliances having computer or processor control including telephones, cellular telephones, televisions, television set top units, point of sale computers, automatic banking machines, lap top computers, servers, personal digital assistants, Smart Cards and automobiles.
The invention will now be described with respect to the particular application of the Data Encryption Standard (DES) encryption and decryption.
Description of DES
The Data Encryption Standard (DES) is a block cipher, where data being encoded or decoded is broken down into sixtyfourbit blocks which are operated upon separately. DES inputs a sixtyfourbit block to be encrypted or decrypted and a sixtyfourbit raw key and outputs a sixtyfourbit result. Only fiftysix bits of the raw key are actually used: the loworder bit of each raw key 8bit byte is discarded, or can be used for parity.
DES will only be described herein with sufficient detail to explain the invention. A more detailed description of (single) DES is provided in FIPS (Federal Information Processing Standards in the United States) publication 463. A description and an extensive discussion are also provided by Bruce Schneier, Applied Cryptography, ISBN 0471117099, John Wiley and Sons, 1996, DES receiving particular attention on pp. 265294.
There are only three kinds of data operations in DES:
The key shift, QPMe at step 44, consists of rotations of the left and right halves of the fiftysixbit key by an identical amount, in a direction and with a number of shift positions determined by the round number and by whether encryption or decryption is being performed. LKP h1h8 at step 50 (performing Sbox substitution) are the eight Sbox lookup tables performed in the round. In the DES standard, the indices for the LKP operations h1h8 at step 50 are each, in effect, preceded by yet another QPM operation, which permutes the six input bits so that the loworder or rightmost bit becomes the bit second from the left in the effective index, but this QPM can be eliminated to match what has been shown above by reordering the elements of the Sbox tables. The Pbox permutation, QPMi at step 52, permutes the results of LKP h1h8 at step 50, presumably to accelerate diffusion of information across all bits.
The XORg operation at step 54 is a simple Boolean exclusive OR on the outputs of the QPMd at step 46 and the output from the QPMf at step 48. Similarly, the XORj operation at step 56 is a simple Boolean exclusive OR on the outputs of the Left In and the output from QPMi at step 52.
Note that all rounds are performed identically except for the previously mentioned differences in the key shift, QPMe, and the swapping of Left Out and Right Out, relative to what is shown in
1.0 Introduction
As noted above, the embodiments of the invention are intended to protect software and data from a whitebox threat model where the attacker has all of the advantages present for an adaptive chosen plaintext attack (control of the number of plaintexts and their content, and access to the resulting ciphertexts), as well as full access to the encrypting software. Thus, the attacker can arbitrarily trace execution and examine all subresults, perform arbitrary static analyses on the software itself, or alter results of subcomputation (e.g., by using breakpoints) to perform perturbation analysis. The only restriction in this model is that the attacker does not have access to the processes by which the executing software was produced.
Hence, whitebox cryptography is aimed at environments where protection is needed but isolation of the cryptographic computation itself from the attacker is for some reason not practical.
Also as noted above, the specific implementation of the algorithm is considered to be irrelevant to security in the blackbox model. In the case of the whitebox model, however, it becomes critical, and changing the specific implementation of the algorithm becomes the primary means for providing security. The invention provides ways to make finding hidden information combinatorially difficult for the attacker.
This description focuses on symmetric block ciphers where encryption requires only substitution boxes and linear transformations, using DES to provide a detailed example of hiding a key in the software. This is done because:
The invention is described with respect to the embedded key case, as opposed to cases where the key is presented dynamically to the software. Clearly though, the invention could be applied to such cases to limit the amount of information available to an attacker.
2.0 Terminology and Notation
This section outlines the terms and notation used in the general description of the invention. A more rigorous treatment is given in the section titled “Additional Observations on the General Invention” which follows.
A bit is an element of the Galois field of the integers modulo 2, that is, the binary set of {0, 1}. A bitvector is a vector over this field, such as the sixbit vector [0 1 0 1 0 0], and a bitmatrix is a matrix over it. Other terms starting with the “bit” prefix are similarly understood.
An expression such as <e_{1}, e_{2}, e_{3}, . . . , e_{k}> is a vector of k elements (the e_{i}'s). Whether the elements are bits will be evident from context.
We denote by P′ an encoded function derived from the function P. _{m} ^{n}P denotes the same function as P, indicating that it maps vectors that are mbits in length, onto vectors that are nbits in length. ^{n}P is simply an abbreviation of _{n} ^{n}P. ^{k}I is the identity function on kvectors. _{m} ^{n}E (a mnemonic for an entropytransference function) is any function from mbit long vectors to nbit long vectors such that, if m≦n, the mapping loses no bits of information, and if m>n, it loses at most n−m bits of information. ^{n}E is, of course, an abbreviation of _{n} ^{n}E. Multiple occurrences of _{m} ^{n}E or ^{n}E in a given formula or equation denote the same function.
v_{i }is the ith element of vector v, and v_{i . . . j }is the subvector containing the ith through jth elements. _{k}v denotes the same bitvector as v, indicating that v has k elements. _{k}e (a mnemonic for an entropyvector) is any vector with k elements. Multiple occurrences of _{k}e within a given formula or equation denote the same vector. x
y is the concatenation of vectors x and y, and x⊕y is the bitwise Boolean exclusiveor (XOR) of x and y.P∘Q is the composition of _{a} ^{b}P and _{c} ^{d}Q where a=d. P∥Q (the concatenation of P and Q) is that function _{a+c} ^{b+d}H such that H(_{a}x
_{c}y)=P(x)∥Q(y), so that H(xy)_{1 . . . b}=P(x) and H(xy)_{b+1 . . . b+d}=Q(y). If P and Q have inverses, then H^{−1}=P^{−1}∥Q^{−1}. The concatenation function “” is associative for vectors, and the concatenation function “∥” is associative for functions.For a matrix M, _{m} ^{n}M denotes M, indicating that M has m columns and n rows. (If we interpret the application of M to a vector as a function application, this notation will be the same as above.)
3.0 DeLinearization and Substitution Boxes
A linear transformation (LT) is a vectortovector transformation function P from vectors to vectors which can be defined by _{m} ^{n}P(_{m}e)=_{m} ^{n}M_{m}e+_{n}d for all _{m}e, where M is a constant matrix and d is a constant displacement vector.
LTs are useful in mixing and reconfiguring information. In the blackbox context, the LTs can be very simple, for example, the Expansion Permutations 46 and Pbox Permutations 52 in DES (see
In the whitebox context, simple LTs cannot be used because they reveal too much information to the attacker. Thus, in the method of the invention, we:
The permutations used in DES are all LTs, as are its bitwise XOR operations 54, 56. Hence, DES is performed by LTs and substitution boxes (SBs) alone.
For a given m and n, there are 2^{mn+n }minput, noutput LTs, but we are primarily interested in those which discard minimal, or nearly minimal, input information (that is, we prefer that m≦n). If m=n, then there are
bijective LTs, since there are
nonsingular n×n matrices (see: Linear Groups, with an Exposition of Galois Field Theory, Leonard E. Dickson, 1958, p. 77, Dover Publications, New York). A nonsingular matrix is a matrix which is invertible.
It is the latter of these two equations which is of greater significance, since we will often use LTs to reconfigure information. Changing the displacement vector, d, of an LT affects only the sense of the output vector elements, and not how the LT redistributes input information to the elements of its output vector. (Recall the general form of an LT as noted above: _{m} ^{n}P(_{m}e)=_{m} ^{n}M_{m}e+_{n}d.)
There are 2^{n}! bijections of the form ^{n}P. Considering the above formulas, we see that the proportion which are linear shrinks rapidly as the value of n increases. For example, there are 2^{3}!=40,320 bijections of the form ^{3}P. Out of this number,
are linear, or exactly one in 30. There are 2^{5}!≈2.631×10^{35 }bijections of the form ^{5}P, of which
are linear, which is about one in 8.222×10^{26}.
Nevertheless, the number of linear bijections ^{n}L becomes very large for large n. For example, there are about 6.442×10^{42}≈2^{142.21 }nonsingular 12×12 matrices. This is a large number, considering that there are only a total of 2^{144 }matrices with dimensions of 12×12.
Note that if A and B are LTs, then so is A∥B, and so is A∘B (where defined). LTs, per se, are of little use in whitebox cryptography, because they are so easily decomposed by Gaussian elimination and related methods.
3.2 DeLinearized LTs and Encoded Functions
Let us consider how we can delinearize LTs and encode functions in general:
3.2.1 Partial Evaluation
Suppose part of our input to a function _{m} ^{n}P is _{b}v, and we wish to replace v with a fixed constant, v=_{b}c. In DES, for example, we have many situations in which we have a transform with two inputs: some data input _{a}x, and a constant encryption or decryption key input _{b}c. By performing partial evaluation, we alter the transform to yield the same output in response to the input _{a}x, but the fixed encryption or decryption key input _{b}c no longer appears in the program.
In mathematical terms, if the input to P is _{a}x
_{b}c, where m=a+b, then we can hide c by replacing P with _{a} ^{n}Q(_{a}e)=P(_{a}ec) for all _{a}e. Thus, we end up with a new transform _{a} ^{n}Q, which performs the same function as _{m} ^{n}P, but incorporates in a concealed way, a constant as part of its input.Note that _{a} ^{n}Q is smaller in size than _{m} ^{n}P. This reduction occurs because part of the input to _{m} ^{n}P is constant; thus we have a restricted use of _{m} ^{n}P, it is not necessary to pass the entire P domain into Q. Therefore, Q is smaller than P but no information is lost.
This process can be implemented in a variety of manners. A simple algorithm for its implementation is presented in the flow chart of
Firstly, for a given function, _{m} ^{n}P, the particular constant input, _{b}c, and its value, must be identified at step 70.
Next, the routine steps through all possible input values, _{a}e, for the original function, _{m} ^{n}P, at step 72. For each possible input value, the output of P(_{a}e
c) is calculated at step 74, and the output of _{a} ^{n}Q(_{a}e) for the given input is set to equal this P(_{a}ec) output at step 76.When it is detected at step 72 that all possible input values have processed, then control passes to step 78 where the P(_{a}e
c) function is replaced in its original program with the new _{a} ^{n}Q(_{a}e) transform, and this particular routine is complete. In the encoded program, only transform Q appears, and there is no transform P or constant _{b}c.In other words, we know what the input is, and we know what output we want to obtain, so we generate a transformation which satisfies those requirements.
Of course, this routine may be incorporated in larger software routines, or be implemented in different manners. As well, transforms P and Q may or may not be linear, and may or may not be matrices.
3.2.2 Encoded Functions and Networks
This technique allows us to encode transforms so that the original transform is incorporated in the encoded program, but cannot be found by attacker. This is useful in DES, for example, where the same transforms are used many times; by encoding a given transform a different way each time it is used, the attacker cannot correlate all the instances of that transform.
For example, for an LT _{m} ^{n}L, we could choose nonlinear bijections ^{m}F_{L }and ^{n}G_{L }such that L′=G_{L}∘L∘F_{L} ^{−1 }is nonlinear. F_{L }is referred to as the input coding, and G_{L }the output coding, of L′ with respect to L, and L′ is an encoded LT. We can similarly encode a nonLT, X, as X′ with input and output codings F_{X }and G_{X }respectively.
Clearly, the encoded algorithm must compensate for the F and G encodings so that it will operate properly. A straightforward way to do this is to implement a network of encodings. For example, if a software algorithm has successive transformations X and Y, we can generate encoded transforms X′ and Y′ which look nothing like X and Y, but in which the output coding of X is corrected by the input coding of Y. Since the output coding of X is incorporated in X′ and the input coding of Y incorporated into Y′, the fact that they are complements is not apparent at all.
A exemplary method of effecting such a process is presented in the flow chart of
Firstly, two adjacent transforms X and Y are identified in the target program at step 90. Next, a random output coding is generated at step 92, for the first transform X (similarly, of course, one could generate a random input coding for the second transform Y).
The encoded transform X′ can now be calculated as X′=G_{X}∘X at step 94. As the input coding to transform Y must compensate for the G_{X }encoding, the encoded transform Y′ can now be calculated as Y′=Y∘G_{X} ^{−1 }(i.e. F_{Y}=G_{X}), at step 96.
The two transforms X and Y can now be replaced in the software program with the encoded transforms X′ and Y′ at step 98, and the routine is complete.
For Y′∘X′ to be an encoding of Y∘X, we require F_{Y}=G_{X}, since Y′∘X′=G_{Y}∘Y∘F_{Y} ^{−1}∘G_{X}∘X∘F_{X} ^{−1}. Variations on this theme can be implemented using “∘”, “∥”, and bijective codings, thus we can construct encoded networks from encoded functions.
These encodings could be implemented a number of ways, for example, the values ^{m}F_{X }and ^{n}G_{X }could be chosen randomly. By being chosen randomly, it would be possible for one of ^{m}F_{X }or ^{n}G_{X }to be linear, but it is probably better to rely completely on randomness than to attempt to avoid linear transforms being generated.
This technique is also useful for disguising sparse matrices. Sparse matrices may provide an attacker with a weakness to exploit. This can be remedied simply by multiplying the sparse matrix with an encoding function which is random but generally dense, and placing the encoded sparse matrix and the inversion of the encoding function into the software algorithm. Thus, rather than facing a sparse matrix, the attacker will face two dense matrices.
3.2.3 Input/Output (I/O)—Blocked Encoding
In section 3.2.2 above, suppose that m and n are inconveniently large when we are encoding some linear or nonlinear function P. The size of DES substitution boxes for example, grow as the power of two of the input size, so an eight bit input will require a table with a manageable 256 entries. However, a 16bit input will require a memory and processing intensive 65,536 entries and 32bits, an impractical 4.3×10^{9 }entries. Therefore, a technique which allows the size of functions to be reduced or divided into multiple, smaller functions, would be very useful.
The growth of inconveniently large functions may easily occur for example, when using the linear blocking method of section 4.1. In that case, we can define input and output codings F_{P }and G_{P }for a function P, using the exemplary routine presented in the flow chart of
First, at step 110, we define m=ja and n=kb, as we want to map _{m} ^{n}P onto _{ja} ^{kb}P. We intend to divide function _{m} ^{n}P into j×k blocks, each block having dimensions of a input bits and b output bits.
Next, we generate two ‘mixing’ linear bijections ^{m}J and ^{n}K (two bijective LTs each of which mixes the entropy of its input bits across all of its output bits as much as possible), at step 112. Alternatively, J and K could simply be randomly generated; whether a certain LT yields the desired mixing depends on the context. The only real restriction on J and K is that they be invertible.
At step 114, we now randomly choose nonlinear input and output coding bijections F and G, which we partition as: ^{a}F_{1}, . . . , ^{a}F_{j }and ^{b}G_{1}, . . . , ^{b}G_{k}. Clearly, there are j input encoding functions F, recognizing that the input to _{m} ^{n}P was divided out as m=ja. Thus, each F encoding function has dimensions a×a. The same applies for the output coding G, which is partitioned into k functions of dimensions: k input bits and k output bits. We then define F_{P}=(F_{1}∥ . . . ∥F_{j})∘J and G_{P}=(G_{1}∥ . . . ∥G_{k})∘K.
We can then generate an encoded function P′=G_{P}∘P∘F_{P} ^{−1 }at step 116 in one of the manners described above: i.e. by traversing the range of inputs to P as described in steps 72 through 78 of
This permits us to connect with a “wide I/O” linear function in encoded form, since, prior to encoding, as a preliminary step, we only need to deal with J and K (that is, we may replace P with K∘P∘J^{−1}), which can be done using the smaller blocking factors of the F_{i}'s and G_{i}'s which we add during encoding.
That is, if the input to P is provided by an LT X, and the output from P is used by an LT Y, we would use J∘X and Y∘K^{−1 }instead. Then the input and output coding of the parts can ignore J and K—they have already been handled—and deal only with the concatenated nonlinear partial I/O encodings F_{1}∥ . . . ∥F_{j }and G_{1}∥ . . . ∥G_{k}, which conform to smaller blocking factors.
As an example of the combinatorics for such encodings, consider the case where we must encode ^{12}P. If we choose a=b=4, then j=k=3, and the number of choices for each of F_{P }and G_{P }is about 6×10^{42 }(nonsingular 12×12 matrices)×9×10^{39 }(choices for sequences of three nonlinear block coding functions): about 5×10^{82}.
This easily extends to nonuniform I/O blocked encoding, i.e. where the groups of input and output encoding function are not all of identical dimensions.
3.2.4 Encoded Function Concatenations
For functions P and Q, one could concatenate them together and then choose an encoding of P∥Q such as G_{P∥Q}∘(P∥Q)∘F^{−1} _{P∥Q }(as per section 3.2.2 above). This mixes the input and output entropy of transform P with that of transform Q, making it harder for an attacker to separate and determine the components P and Q.
3.2.5 ByPass Encoding
There are instances in which it is necessary to execute a transform, say _{m} ^{n}P, but also to pass some or all of the input bits m, onto a subsequent operation. In DES, for example, there are a series of steps in which the data input on the righthand side becomes the data for the lefthand side for the next round (see
In general, this is done by concatenating a new function to the original transform, which does the job of transferring the desired bits through the encoded transform without being affected by the original transform. To carry a extra bits of entropy at the input, and b extra bits of entropy at the output, of _{m} ^{n}P, where a≧b, we can encode _{m+a} ^{n+b}P′ as G_{P∥E}∘(P∥_{a} ^{b}E)∘F^{−1 } _{P∥E}, where _{a} ^{b }E is the bypass component of P′.
The process of generating the transform _{m+a} ^{n+b}P′ is basically the same as described above with respect to the generation of other encoded transformations. As per the flow chart of
Input and output encoding transforms F and G are then generated at step 132, in the manner described herein.
The encoded transform is then generated at step 134, in a manner consistent with the nature of the transform, for example, by traversing all possible inputs to the original transform _{m} ^{n}P, and defining _{m+a} ^{n+b}P′ as equal to G_{P∥} _{ a } _{b E}∘(P∥_{a} ^{b}E)∘F_{P∥} _{ a } _{b E} ^{−1}. If the original transform _{m} ^{n}P is a matrix, for example, then _{a} ^{b}E will be a matrix that selects the desired bits from the input, and allows them to be passed through the encoded transform _{m+a} ^{n+b}P′.
Next, we compensate for the input and output encoding transforms F and G (if necessary) by generating new compositions for transforms adjacent to the original transform _{m} ^{n}P in the targeted software algorithm, at step 136 (say transforms X and Y). We then replace original transform _{m} ^{n}P, and adjacent transforms X and Y, in the targeted software program, with their respective encoded transforms P′, X′ and Y′ at step 138.
If m+a=n+b, then P′ is locally secure. A function which is locally secure has the property that every original function is just as likely as every other original function. Thus, given only the encoded results, an attacker cannot tell with any probability whether one original input was more likely than another (see section 3.3).
If _{a} ^{b}E has the specific form ^{a}I (so that a=b), it can be referred to as identity bypass encoding, so the compensation at step 136 will not have to be performed. Nonidentity bypasses may also be performed, but the nonidentity transformation encoding will have to be undone. This can easily be done by incorporating the inversion transform into the adjacent transforms at step 136.
However, varying the input and output codings arbitrarily, if a>b, we cannot arbitrarily substitute the bypass component without affecting the set of functions obtained. Its (vectorset valued) inverse mapping implies a particular partition of the possible input subvectors for that component (where elements of a partition element agree on the original P input vector). We cannot entirely hide the cardinalities (i.e. the number of elements) in the subsets forming this partition.
That said, even if some statistical attack finds these cardinalities, it may yet remain quite difficult to resolve the input and output codings of P′.
3.2.6 SplitPath Encoding
For a function _{m} ^{n}P, we may disguise it by concatenating it with another function which may be arbitrary. For example, we could encode function _{m} ^{n}P into function _{m} ^{n+k}X, where X(_{m}e)=P(_{m}e)
_{m} ^{k}R(_{m}e) for all _{m}e, and R is some arbitrary, fixed function.The effect is that, if P is lossy (i.e. information can be obtained by observing processing on transform P), then X may lose less information or no information at all (processing on X is obscured to the observing, by the addition of the processing on R). In particular, we can use this method to achieve local security, as described in section 3.3.
The generation of the encoded transform X, is done in the same manner described above with respect to
This technique adds new redundant processing into the data flow of the program. This redundant processing is substantive in that it becomes integral with the encoded transform X, and cannot easily be identified or removed. This is in contrast to “dummy code” used in the art, which does not actually execute and can be identified using data flow analysis techniques. With the method of the invention, it is quite difficult to distinguish which operations are from the original software code, and which have been added.
As these redundant arguments ultimately have no impact on the outcome of the software program, they can be generated randomly. Techniques for generating random and pseudorandom numbers are known in the art.
3.2.7 Simultaneous ByPass Encoding
In some cases, we may wish to perform some function on an input, and yet preserve the input in the output for later use (but not separate from the output as we did in section 3.2.5). This may be feasible if we can arrange that our function is encoded as a bijection, as a bijection has a one to one mapping between inputs and outputs.
Suppose, for example, that we have a bijective encoded function P′=G_{X}∘X∘F_{X} ^{−1 }where X is derived from P using splitpath encoding (see section 3.2.6). We can view this same function as Q′=G_{Q}∘Q∘F_{Q} ^{−1 }where we define Q=F_{Q}=^{m}I for an appropriate m, and G_{Q}=P′. That is, we use an identity input encoding for an identity and regard P′ as the output coding. Then applying its inverse, P′^{−1}, to the output, we retrieve the input. By encoding the inverse of P′ (for example, combining it with a subsequent transformation), we can conceal the use of the inverse of P′, which need not appear as such in an implementation.
The generation of the encoded transform is done in the same manner described above with respect to
3.2.8 Output Splitting
This technique is useful for disguising outputs where input information can be well hidden. This does not appear to be the case for DES: for implementations of DES, output splitting is not recommended since it cannot provide much security.
Where the technique is appropriate, to make statistical “bucketing” attacks more difficult, we may encode a function P as k≧2 functions, P_{1}, P_{2}, . . . , P_{k}, where each encoding can mix in additional entropy as described in sections 3.2.4 or 3.2.5 above, and where the output of all of the encoded P_{i}'s is needed to determine the original output of P.
For example, given a function _{m} ^{n}P, we can choose k=2, define _{m} ^{n}P_{i }to be a randomly selected fixed _{m} ^{n}E, and define _{m} ^{n}P_{2 }(_{m}e)=P(_{m}e)⊕P_{1 }(_{m}e) for all _{m}e.
At this point, we can compute the P output from the exclusiveor of the outputs of the two P_{i}'s. However, after we then independently encode the P_{i}'s, the output of P_{1 }and P_{2 }is not combinable via an LT into information about P's output.
3.3 Substitution Boxes and Local Security
We can represent any function _{m} ^{n}P by a substitution box (SB): an array of 2^{m }entries, each of n bits. To compute P(x), find the array entry indexed by the binary magnitude x. The exponential growth in SB size with its input width, limits SBs to the representation of narrow input functions.
When an SB represents L′ or X′ (where L is an LT and X is a nonlinear transform), then if m=n, and L or X is bijective, the SB for L′ or X′ bijection is locally secure: irrespective of L or X, we can always choose an input or output coding such that the SB of L′ or X′ is that SB. Locally, any information about L or X, other than the value of n and that it is a bijection, cannot possibly be discovered from the encoded SB of L′ or X′ itself (which only means, of course, that an attack must be nonlocal).
The lossy case is not locally secure. When a slightly lossy encoded function is represented as an SB, some information about the function beyond its input and output widths can be found by examining its SB. Completely understanding it, however, still requires a nonlocal attack (as we will see in the DES example).
The price of using such a protean representation is that SBs are often redundant, and may include considerable nonproductive entropy. For some practical implementations, one may need something less bulky. Of course, any substantial implementation shrinkage compared to an SB implies restrictions on input and output codings.
Plainly, for any function _{m} ^{n}P and any bijection _{m}B, the SB for P∘B has the same elements as the SB for P, but (unless B=^{m}I) in a different order.
4.0 WideInput Encoded LTs: Building Encoded Networks
Because of the previously noted immense storage needs of wideinput SBs, it is infeasible to represent a wideinput encoded LT by an SB. One can, however, construct networks of SBs which implement a wideinput encoded LT.
4.1 A Blocking Method
The following construction is protean: it can handle LTs in considerable generality, including compositions of LTs, and for a wide variety of LTs of the form _{m} ^{n}L encoded as _{m} ^{n}L′, the form of the network can remain invariant except for variations in the bit patterns within its SBs.
For an LT, L, we simply partition the matrix and vectors used in the LT into blocks, giving us wellknown formulas using the blocks from the partition which subdivide the computation of L. We can then encode the functions defined by the blocks, and combine the result into a network, using the methods in section 3.2 above, so that the resulting network is an encoding of L.
Consider an LT, L, defined by _{m} ^{n}L(_{m}e)=_{m} ^{n}M _{m}e+_{n}d for all _{m}e: we choose partition counts m_{#} and n_{#} and sequences <m_{1}, . . . , m_{m} _{ # }> and <n_{1}, . . . , n_{n} _{ # }>, such that
That is, the former sequence (the mpartition) is an additive partition of m, and the latter sequence (the npartition) is an additive partition of n.
The mpartition partitions the inputs and the columns of M; the npartition partitions d and the outputs. Hence, the i, jth block in partitioned M contains m_{i }columns and n_{j }rows, the ith partition of the input contains m_{1 }elements, and the jth partition of d or the output contains n_{j }elements.
At this point, it is straightforward to encode the components (of the network forming L) to obtain an encoded network, by the methods of section 3.2, and then representing it as a network of SBs (see section 3.3). In such a network, none of the subcomputations is linear each is encoded and represented as a nonlinear SB.
A naive version of this consists of a forest of n_{#} trees of binary ‘vector add’ SBs, with m_{#} (m_{#}−1) ‘vector add’ nodes per tree. At the leaves are m_{#} unary ‘constant vector multiply’ nodes, and at the root is either a binary ‘vector add’ node (if there is no displacement) or a unary ‘constant vector add’ node (if there is a displacement).
However, we can eliminate the unary ‘constant vector add’ and ‘constant vector multiply’ nodes entirely. We simply compose them into their adjacent binary ‘vector add’ nodes, thereby saving some space by eliminating their SBs.
A potential weakness of this entire approach is that the blocking of L may produce blocks, such as zero blocks, which convert to SBs whose output contains none, or little, of their input information. This narrows the search space for an attacker seeking to determine the underlying LT from the content and behaviour of the network. However, so far as we have yet determined, such blocked implementations remain combinatorially quite difficult to crack, especially if we apply the proposals below.
For example, the potential weakness noted above, can be addressed as follows:
While the above methods help, it is not easy, in general, to eliminate m×n blocks which lose more bits of input information than the minimum indicated by m and n. For example, if we partition a matrix ^{kn}M into k×k blocks, we cannot guarantee that all of the k×k blocks are nonsingular (i.e. can are invertible), even if the rank of M is greater than k. Hence, if M is nonsingular, a partition of M into square blocks may contain some singular (lossy) blocks.
Therefore, some information about an encoded LT may leak in its representation as a blocked and delinearized network of SBs when this blocking method is used.
5.0 Example: An Embedded, Hidden Key Implementation of DES
We now discuss an implementation of DES that is resistant to a whitebox attack. The implementation presented here has weaknesses, both in security and efficiency, which are addressed in sections 5.4 and 7.
As described above with respect to
5.1 Replacing the DES SBs
In
K_{r }is the 48bit subkey for a given round r, which is constant for a particular application. As noted above, DES only uses 48 bits of the 56 bit key in a given round.
This section describes how we replace the DSBs with new SBs so that:
In each round, a DSB's input is the Boolean exclusiveOR (XOR) of ‘unpredictable’ information, not determined by the algorithm plus the key, and ‘predictable’ information, determined by the algorithm and the key. This predictable information can be determined in advance, without any knowledge of the unpredictable information.
Hence, we can dispense with the ‘predictable’ information entirely by modifying the DSBs into new SBs. The reason is that the XOR of the ‘unpredictable’ information (the argument) with ‘predictable’ information (a constant) is a bijection (see the last paragraph in section 3.3).
Let us therefore produce new SBs identified as _{K} ^{r}S_{i}, where K is the encryption key, r is the round number, and i is the corresponding DSB number, such that, for any given input, _{K} ^{r}S_{i }yields the same result as S_{i }would produce in round r if the DES key were K, but the XORs of the inputs of the original DSBs have been eliminated (see section 3.2.1). Each of the _{K} ^{r}S_{i }substitution boxes, of which there are 16×8=128 (16 rounds, 8 Sboxes per round), is still in _{6} ^{4}E form.
At this point, the overt key K has disappeared from the algorithm: it is represented in the contents of the _{K} ^{r}S_{i}'s. This permits us to remove the XORs (“⊕”) with the inputs to S_{1}, . . . , S_{8 }shown in
5.1.2 Preparing the Modified DSBs for Local Security
In greybox (Smart Card) implementations of DES the DSBs have proven to be effective sites for statistical bucketing attacks (again, see: Introduction to differential power analysis and related attacks, 1998 by Paul Kocher, Joshua Jaffe and Benjamin Jun). To make such attacks more difficult for our whitebox implementation, we replace the _{K} ^{r}S_{i}'s with new SBs which can be made locally secure (see section 3.3) using suitable input and output encodings. That is, we convert the SBs into ^{8}E form by defining:
_{K} ^{r} T _{i}(_{8} e)=_{K} ^{r} S _{i}(_{8} e _{1 . . . 6})∥R(_{8} e)
for all _{8}e, for the fixed key K, for round r=1, . . . , 16, for SB number i=1, . . . , 8, where we define R(_{8}e)=<_{8}e_{1}, _{8}e_{6}, _{8}e_{7}, _{8}e_{8}> for all _{8}e. This is an application of splitpath encoding (see section 3.2.6).
The plan is that the first six bits of the input of a transform _{K} ^{r}T_{i }will be the 6bit input to DSB i in round r. We then add two extra input bits. Hence, the left 4bit half of the output of transform _{K} ^{r}T_{i }is the output of DSB i in round r, and the right 4bit half contains the first and last input bits of DSB i in round r followed by the two extra input bits. That is, the right half of the output contains copies of four of the input bits.
To see that each transform _{K} ^{r}T_{i }is a bijection, we note that the function F_{a, b, c, d}defined for any constant bits a, b, c and d by F_{a, b, c, d}(_{4}e)=_{K} ^{r}T_{i}(<a>∥_{4}e∥<b,c,d>) is a bijection. Every row of every DSB contains a permutation of <0, . . . , 15>, with the row selected by the bits corresponding to a, b above. The XOR with the relevant bits of the key K effectively reorders this permutation into a new permutation. The output of F_{a, b, c, d }is therefore a bijection, mapping the _{4}e according to a 1 to1 mapping of the input space determined by a permutation. Since _{K} ^{r}T_{i }simply copies the bits corresponding to a, b, c and d to the output, transform _{K} ^{r}T_{i }preserves all of its input entropy; that is, it is a bijection.
5.1.3 Providing 64 bits of ByPass Capacity
At this point, in order to carry all data from each round into the next, our first thought would be to employ simultaneous bypass (see section 3.2.7), whereby the eight _{K} ^{r}T_{i}'s in round r would have sufficient capacity to carry all of the information from a previous round into the next round, and each _{K} ^{r}T_{i }would be locally secure (see section 3.3). The capacity needed is 64 bits, and 8×8=64.
Unfortunately, there are two problems with this idea:
Since each _{K} ^{r}T_{i }must be a bijection to support local security, and we will not use 8×8 decoding, we are therefore prevented from using simultaneous bypass encoding. As a result, each _{K} ^{r}T_{i }effectively carries only four bits of input to the next round.) This includes two bits from the right side data path of DES, plus the two extra input bits which we can take from wherever we wish. The bypass capacity of the _{K} ^{r}T_{i}'s is too small by 32 bits.
So we add four more SBs, designated (preencoding) as _{K} ^{r}T_{9}, . . . , _{K} ^{r}T_{12}. Each is a linear instance of ^{8}E prior to delinearization and encoding. They provide the remaining 32 bits: 16 bits of rightside bypass capacity, and 16 bits of leftside bypass capacity.
5.2 Connecting and Encoding the New SBs to Implement DES
The overall dataflow structure of our DES implementation immediately prior to delinearization of LTs and encoding of SBs (see sections 3.2 and 3.3), is shown in the block diagram of
The dataflow would look just the same after delinearization and encoding, except that each M_{i }would be replaced by a corresponding encoded transform M_{i}′ and each _{K} ^{r}T_{i }would be replaced by a corresponding _{K} ^{r}T′_{i}. Except for the fact that the transforms have been encoded (represented by the “′” characters) it would be identical.
5.2.1 DataFlow and Algorithm
Before delinearization and encoding, each M_{i }is representable as a matrix, with forms _{64} ^{96}M_{1}, _{96} ^{96}M_{2}, and _{96} ^{64}M_{3}, respectively. These transforms and how they are generated are discussed in section 5.2.2.
In
The appearance of rows of _{K} ^{r}T_{i}'s in order by i in
Without the “′” characters (that is, prior to delinearization and function encoding: see section 3.2) the modified DES algorithm may be represented in pseudo code as follows:
v=M_{1(intext)}
for r:=1, . . . , 16loop
v:=M _{2}(_{K} ^{r} T(v))
end loop
outtext:=M _{3}(v)
We use the terms intext and outtext in the pseudo code, rather than plaintext and ciphertext, because:
The algorithm is the same, except for addition of the “′” characters, after delinearization and function encoding.
5.2.2 The Transfer Functions
In constructing M_{1}, M_{2}, and M_{3}, we must deal with the sparseness of the matrices for the LTs used in standard DES. The bitreorganizations, such as the Expansion 46 and Pbox transform 52 appearing in
Therefore, we use the second method proposed for handling sparseness in section 4.1: doubling the implementations into two blocked implementations, with the initial portion of each pair being a ‘mixing’ bijection. We will regard this as part of the encoding process, and discuss the nature of the M_{i}'s prior to this ‘antisparseness’ treatment.
The following constructions all involve only various combinations, compositions, simple reorganizations, and concatenations of LTs, and are therefore straightforward:
We recommend using 4×4 blocking for the M_{i}'s. As a result of the optimization noted in section 4.1 this means that the entire implementation consists entirely of networked 8×4 (“vector add”) and 8×8 (_{K} ^{r}T_{l}′) SBs.
Aside from M_{1}'s input coding and M_{3}'s output coding, both of which are simply _{64}I (appropriately blocked), all SBs are input and outputcoded using the method of section 3.2.3 in order to match the 4bit blocking factor required for each input by the binary ‘vector add’ SBs.
5.2.4 Naive and Recommended Variants
The above material completes the naive or ‘naked’ variant of whitebox DES. The recommended variant pre and postwhitens DES as follows: we use an intext and outtext encoded as a whole using the method of section 3.2.3 (or better yet, similarly encoded but in much larger blocks, with chaining). This completely foils statistical ‘bucketing’ attacks depending on the control of unencoded bits.
That is, we modify the scheme shown in
The additional mixing transforms M_{0 }and M_{4 }can be added in separate processing steps after the above is performed.
5.3 Complexity of Attacks on the Naive Variant
In its naive form, where intext is plaintext (for encryption) or ciphertext (for decryption, or outtext is ciphertext (for encryption) or plaintext (for decryption), the DES implementation can be cracked by statistical bucketing. Thus, one should generally avoid the naive form for DES applications. Instead, one should use an encoded intext and an encoded outtext.
The attacker cannot possibly extract information from the _{K} ^{r}T_{i}′ transforms themselves as they are locally secure (see section 3.3).
By far the best place to attack our implementation in naive form seems to be at points where information from the first and last rounds is available. In the first round (round 1), the initial input is known (the M_{1 }input is not coded), and in the last round (round 16), the final output is known (the M_{3 }output is not coded).
Attacks should be focussed on the first (1) and final (16) rounds. Cracking either round 1 or round 16 provides 48 key bits; the remaining 8 bits of the 56bit DES key can then be found by bruteforce search on the 256 remaining possibilities using a reference DES implementation.
For an attack on round 1, the attacker should focus on the _{K} ^{r}T_{i}′ inputs to round 2, and for an attack on the last round, on the _{K} ^{r}T_{i}′ outputs from round 15. The attacker then deals with the input or output information after it has been broken up from (round 1), or before it has been merged into (round 16), the 8bit bundles input to and output from the _{K} ^{r}T_{i}′ transforms.
Consider any round 1 DSB S_{i }in standard DES. Its input is affected by 6 bits of the round 1 subkey, and two of its output bits affect only one DSB in round 2. We focus on one of these output bits, which we denote b. S_{j }will then refer to the round 2 DSB which is affected by b.
We make a guess on the 6 bits of subkey affecting S_{i}, and choose a set of plaintexts, denoted I_{0}, which have the property that if our key guess is correct, bit b will have a value of 0 for the encryption of each plaintext in the set. We similarly choose a set I_{1}, such that if our key guess is correct, b will have a value of 1.
The encryption of the texts in I_{0 }(resp. I_{1}) will lead to a set of inputs I_{0}′ (resp. I_{1}′) to ^{2}T_{z} _{ j }. The important point is that if our key guess is correct, I_{0}′ and I_{1}′ must necessarily be disjoint sets, whereas if our key guess is incorrect, the elements of I_{0}′ and I_{1}′ will be instances of _{8}e. We would like the probability that no collision (an element occurring in both I_{0}′ and I_{1}′) occurs in the event of an incorrect key guess to be no greater than 2_{−6}. Experimentally, this occurs when I_{0}′=I_{1}′≈27.
We note that DSB output bits {3, 7, 11, 15, 18, 24, 28, 30} have the property that each bit comes from a unique SB and goes to a unique SB in the following round. By tracking these bits, we can search for the subkey affecting each round 1 DSB in parallel (this requires a clever choice of elements for I_{0 }and I_{1}, because of the overlap in the inputs to the round 1 DSBs). Again, experimentation shows that fewer than 2^{7 }plaintexts are necessary in total to identify a very small set of candidates for the 48bit round 1 subkey. The remaining 8 bits of key can subsequently be determined by exhaustive search.
This gives a cracking complexity of 128 (chosen plaintexts)×64 (number of 6 bit subsubkeys)+256 (remaining 8 bits of key)+256 encryptions. This attack has been implemented, and it successfully finds the key in under 10 seconds.
Note of course that this represents a worstcase scenario for a particular DES application. The effectiveness of each method of the invention will depend on the particulars of the specific application, which can easily be analysed.
5.4 Complexity of Attacks on the recommended Variant
The recommended variant of the implementation uses an intext and outtext encoded as a whole using the method of section 3.2.3 (or better yet, similarly encoded but in much larger blocks, with chaining). This completely foils statistical ‘bucketing’ attacks depending on the control of unencoded bits.
That is, we modify the scheme shown in
In other words, the “recommended variant” effectively makes the input and output unknown to the attacker by prepending and appending what are, in effect, random ciphers. As a result, the attackers expectations, based on the original cipher, are not met, and the normal statistical approach to an adaptive chosen plaintext attack is foiled. It would be necessary for the attacker to crack:
We do not know how complex an attack on this variant is, since we have not yet found an effective way to attack it. The difficulty of cracking the individual encodings suggests that it will have a high complexity.
The weakest point would seem to be the blockencoded wideinput LTs. However, it is not merely a matter of finding weak 4×4 blocks (ones where an output's entropy is reduced to three bits, say, where there are only 38,976 possible nonlinear encodings). The first problem is that the output will often depend on multiple such blocks, which will then require some power of 38,976 tries. Of course, as previously noted, we may guess part of such encodings.
However, we must still deal with the second, and much more difficult, problem, which is: once the attacker has a guess at a set of encodings, partial or otherwise, for certain SBs, how can it be verified? Unless there is some way to verify a guess, such an attack cannot be effective.
6.0 How Useful is it?
It may not be immediately clear that the recommended variant (see section 5.4) of our whitebox DES implementation, or whitebox implementations of other ciphers using the recommended variant, can be useful. We consider here how such implementations can be made to do useful work.
6.1 Handling Plain Input or Output
Although we noted that the recommended variant employed an encoded intext and outtext, we can, in fact, employ an unencoded intext or outtext. After all, we are computing an invertible function; adding further bijections as pre or postwhitening will not change this fact, and the result is that we are computing a cipher which embeds DES, rather than DES itself.
As a result, any obvious kind of statistical “bucketing” attack based on properties of DES is foiled.
The price we pay (other than the current slowness and size of whitebox implementations) is that we are no longer using a standard encryption algorithm. However, it seems reasonably likely to be as strong as the original algorithm under blackbox attack, and is certain to be much stronger under whitebox attack.
6.2 WhiteBox Transaction Processing
From the description herein, it is straightforward to add some decision capability and flowcontrol and, indeed, to produce a Turingcomplete set of operations (i.e. an operation set whose elements can be combined to compute anything which is computable), which nevertheless can be implemented in an inputoutput encoded fashion similar to our SBs. (The building blocks might be SBs, and networks of SBs, with added interpretive behavioural capabilities.) We can use this to perform fileupdates and the like, in such a fashion that we decrypt encoded data to encoded, but usable, form, modify the information, encrypt it to encrypted and encoded form, and store it again.
So long as only small amounts of information enter or leave a transaction in plain form at unsecured sites, and almost all of the information transferred is encoded and encrypted for storage and at least encoded for update at unsecured sites, we can then perform file updates and transaction processing using whitebox cryptography in such a fashion that nothing substantial ever leaves the encoded world, although some data at times is not in the encrypted world, except at secured sites. Any substantial decrypting and decoding of information can then be reserved for sites which are very well controlled.
This provides a means whereby software can be protected against insider attacks, particularly at sites which cannot otherwise be well protected by their owners. For example, if we fear military or industrial espionage, transaction processing as sketched above might be a significant help in managing that risk.
6.3 WhiteBox “Whitening”
It is sometimes recommended to use “pre and post whitening” in encryption or decryption, as in DESX (for a description, see: Applied Cryptography, Bruce Schneier, 1996, 2nd ed., pp. 265301, John Wiley and Sons, New York). We note that the recommended variant of the invention per section 5.2.4 computes some cipher, based on the cipher from which it was derived, but the variant is quite an nonobvious one. In effect, it can serve as a form of “super pre and postwhitening”.
In effect, it allows us to derive innumerable new ciphers from a base cipher. All attacks on cryptography depend on some notion of the search space of functions which the cipher might compute. The whitebox approach increases the search space, generally by a significant amount.
6.4 WhiteBox Asymmetry and WaterMark
The effect of using the recommended variant is to convert a symmetric cipher into a oneway engine: possession of the means to encrypt in no way implies the capability to decrypt, and vice versa.
This means that we can give out very specific communication capabilities to control communication patterns by giving out specific encryption and decryption engines to particular parties. By using double encryption or decryption based on a pass phrase, we can arrange that changing the communication patterns requires both a communicated pass phrase and a communicated encryption or decryption engine. And of course, every such engine is effectively watermarked by the function it computes.
It is also possible to identify a piece of information by the fact that a particular decryption engine decrypts it to a known form. There are many variations on this theme.
7.0 Logical Extensions
Many improvements, as well as extensions to other kinds of cryptographic computations, follow logically from the teachings herein. For example, there are clearly improvements that can be made in applying this work to RSAlike computations.
For DESlike ciphers, the blocking technique in section 4.1, while general, is of O(m_{#} ^{2}n_{#}) complexity for a fixed bound on SB size. It is clear that other network topologies have potential to improve on the techniques described herein, such as shuffleexchange topologies (especially irregular ones, for security). Where m=n, for example, these have the potential to reduce the spatial complexity for representing LTs to O(m_{#} log n_{#}) for a fixed bound on SB size.
The methods of the invention can also be combined with general methods of tamperproofing software or applied to software other than cryptographic computation software.
Additional Observations on the General Invention
This section contains comments on the techniques described above regarding whitebox cryptography, and alternative approaches.
8.0 Potential Avenues of Attack on Obfuscated DES Implementations
The following properties of the hiddenkey DES implementation described in the copending patent application filed under the Patent Cooperation Treaty (PCT), as serial number PCT/CA00/00677, titled: “Tamper Resistant Software Encoding”, are potential avenues for attack:
A function f is information preserving precisely if it is an injection (is ‘one to one’); that is, f(x)=f(y) implies that x=y. We are primarily interested in information preserving Boolean functions.
9.2 Linear Boolean Functions from m Inputs to n Outputs
To define a concept of linearity, we need to embed Boolean vectors in a vector space. To make a vector space possible, we need a Boolean field. As described in more general terms above, we use the field F_{2}=<Z_{2}, +, *> where Z_{2 }is the integers modulo 2 (that is, the set {0, 1}), + is integer addition modulo 2, (which has the same effect as Boolean XOR), and * is integer multiplication (which has the same effect as Boolean AND).
We define a vector space V_{n}(F) over a field F to have elements in F^{n }(denoting the ordern Cartesian product of F with itself). We define the product of multiplication of a vector by a scalar as the vector in which each new element is the product of the original element and the scalar in F_{2}. We define vector addition as elementwise addition of corresponding elements.
A linear nton Boolean function F is information preserving if, when F is defined by Y=F(X)=AX+B as described above, A's determinant is nonzero. Since we are in F_{2}, that means that its determinant must be 1.
There are 2^{(n} ^{ 2 } ^{+n) }linear nton Boolean functions, since there are n^{2 }slots in A and n slots in B to fill independently with either 0 or 1.
For the case of an mton Boolean function f defined by F(X)=AX+B, F is onetoone if and only if n is not less than m, and the set of m columns of A, as a set of Boolean vectors, is the basis of a vector space of dimension m (that is, if the columns as Boolean vectors are linearly independent).
9.2.1 Counting Linear Functions
As noted above, there are 2^{mn+n }linear functions with m inputs and n outputs. For the case of m=n, where no information is lost, there are therefore 2^{n(n+1) }linear functions and there are
nonsingular n×n matrices over F_{2}.
However, a linear function also includes a displacement vector; for each of the above matrices, we can choose any of 2^{n }displacement vectors. Hence, the total number of n×n linear bijections (that is, linear permutations) is
For example, the number of 2×2 linear permutations is:
2^{2}(2^{2}−2^{0}) (2^{2}−2^{1})=4×3×2=24
Since there are only 24 permutations of the numbers 0 . . . 3 in total, all 2×2 permutations are linear.
An n×n linear function in F_{2 }is singular if and only if, of the n column vectors (or equivalently, of the n row vectors) of its matrix, either at least one is zero, or a subset of the nonzero vectors yield a zero vector when its members are all added together. That is, either there is a zero vector, or there is a subset S of the vectors such that each member is nonzero and each vector bit position is 1 in an even number of the members of S. (For example, the vectors <0, 1, 1>, <1, 0, 1>, and <1, 1, 0> form a linearly dependent set since their sum is <0, 0, 0>.)
To restate the above, an n×n linear function is singular in F_{2 }if and only if there is a nonempty subset S, not necessarily proper, of the column vectors, such that each bit position is 1 in an even number of the members of S.
If a set S of vectors is linearly dependent, then any set of vectors which contains S is also linearly dependent.
9.3 General Boolean Functions from m Inputs to n Outputs
If we consider unrestricted Boolean functions from m inputs to n outputs, there are 2^{(2} ^{ m } ^{n) }of them. (There are 2^{m }possible input vectors. Corresponding to each, there is a vector of n elements. That gives a total of 2^{m }n slots to fill independently with either 0 or 1. Any difference in any bit produces a different function, so there are 2^{(2} ^{ m } ^{n) }functions.)
Such a function can only be informationpreserving (an injection, onetoone) if n is greater than or equal to m.
9.4 Permutations
If we restrict our attention to cases of nton Boolean functions, of the 2^{(2} ^{ n } ^{n) }of them, (2^{n})! are permutations of the set of all Boolean vectors of length n (that is, in effect, of the numbers from 0 to 2^{n}−1). All permutations are informationpreserving (indeed, any injection, and hence, bijection, from a totally ordered set to itself can always be viewed as a ‘permutation’, and is commonly so called in cryptology), but not all are nonlinear. For example, neither the identity permutation, nor the permutation which inverts every input bit, are nonlinear. In the former, for F(X)=AX+B, A is the identity matrix and B is all zeros. In the latter, A is the identity matrix and B is all ones.
If we consider 2to2 Boolean functions, (2^{2})!=24 of them are permutations. All of the permutations are linear, which can be seen as follows:
In F(X)=AX+B form, there are six nonsingular choices for A. For each of these, there are four possible choices for B, so 6*4=24. Plainly, for each choice of B, the effect of B is to perform a permutation, since it at most flips a bit position. Hence, for AX+B to yield a permutation of X, it is sufficient for AX to yield a permutation of X, but the fact that A yields a 1to1 function guarantees that it is a permutation. It follows that the 24 permutations of all Boolean vectors of length 2 are identical to the 24 2to2 linear bijections in F_{2}.
If we consider 3to3 Boolean functions, (2^{3})!=40,320 of them are permutations of the numbers from 0 to 7. However, there are only:
2^{(3} ^{ 2 } ^{+3)}=2^{12}=4,096
linear 3to3 Boolean functions in total (ignoring information preservation). It follows that most of the permutations of all possible Boolean vectors of length 3 are nonlinear.
The number of linear 3×3 Boolean functions which are permutations is:
2^{3}(2^{3}−2^{0})(2^{3}−2^{1})(b^{3}−2^{2}) =8×7×6×4 =1,344.
The exact number of nonlinear 3×3 Boolean permutations is therefore:
40,320 permutations−1,344 linear permutations =38,976.
A choice among them therefore contains about 15.25 bits of entropy. “Bits of entropy” are bits of arbitrary or randomly chosen information. Wherever a given string can have K distinct equiprobable values, then we find the number of bits of information, or entropy, in K, by solving the equation 2^{b}=K for b, i.e. b=(log K)/(log 2). In other words, b is the number of bits needed to encode K distinct choices. Referring to the above example, b=(log 38,976)/(log 2)≈15.25.
10.0 Adding Extra NonLinearity to Functions Such as DES
To reduce the added entropy and hence, the size of the subnetworks which provide it, we should favour permutations of 3bit noncontiguous bitgroups. Each represents a number from 0 to 7, so there are 8 elements and 8!=40,320 possible permutations. However, 3 does not necessarily evenly divide the number of bits for which we need a nonlinear unary operator.
10.1 Building Wider Permutations from Narrower Ones
If we take the 48 bits coming out of the XOR 54 with the round subkey (see
However, if we take the 32 bits coming out of a round (see the output of XOR 56 in
In the above, we have 1bit overlap to produce a 5×5 permutation. Using an overlap of 2 bits as shown in the block diagram of
In the case of the 5×5 permutation produced as above, if we restrict each 3×3 subpermutation to be nonlinear, we have 38,976^{2}/2=759,564,288 of them. (There are 38,976 choices for each, but there are 2 ways to map two things onto two things for the input bit in the 2nd which is from the 1st.)
This gives about 29.50 bits of entropy for a choice among them.
If, instead, we were to use a nonlinear 5×5 permutation directly, then we have 32!=2.631×10^{35 }permutations in all. Of these, the number of linear permutations is:
2^{5}(2^{5}−2^{0})(2^{5}−2^{1})(2^{5}−2^{2})(2^{5}−2^{3})(2^{5}−2^{4}) =32×31×30×28×24×16 =319,979,520 =3.199×10^{8}
Hence, the linear permutations are a very small fraction of the total, 3.199×10^{8 }of 2.631×10^{35 }permutations.
A choice among the nonlinear 5×5 permutations therefore contains about 117.66 bits of entropy, so using the above construction loses about 88.16 bits of entropy, but saves storage space. (Of course, we can apply this same technique with higher numbers; e.g., building a 7×7 permutation from two 4×4 permutations.)
Similarly, if we use the overlapping of
If, instead, we were to use a nonlinear 4×4 permutation directly, then we have 16!=2.092×10^{13 }permutations in all. Of these, the number of linear permutations is:
2^{4}(2^{4}−2^{0})(2^{4}−2^{1})(2^{4}−2^{2})(2^{4}−2^{3}) =16×15×14×12×8 =322,560.
Again, the linear permutations are a very small fraction of the total.
A choice among the nonlinear 4×4 permutations thus contains about 44.25 bits of entropy. So the construction loses about 18.33 bits of entropy, but takes less storage space.
Then to do 64 bits, we use twenty 3×3 permutations and one 4×4. To do 32 bits, we do nine 3×3 permutations and one 5×5 permutation.
The idea is to slice DES so that we have graph cuts at which we compute a function from M bits to N bits. The most obvious case of this is a round pair, which computes a function from 64 bits to 64 bits.
(Each round of DES simply copies the right 32bit half of its 64bit input to the left 32bit half of its 64bit output, while placing a mixture of the left and right input halves in the right 32bit half of its 64bit output. Accordingly, to mix both halves, we have to perform two rounds; one round is insufficient. A “round pair” is typically an oddnumbered round followed by an evennumbered round: 1 and 2, or 5 and 6, or 11 and 12, or 15 and 16.)
We represent the k'th 64bit slice S_{k }bitvector by another 64 element bitvector, M_{k}(S_{k}). Where the old roundpair function was F_{k}, applied as in:
S _{k+1} =F _{k}(S _{k})
the new round function is:
M_{k+1}∘F_{k}∘M_{k} ^{−1}
where we seek to make the composition highly nonobvious and hard to unravel into the above compositional form.
10.2 Banyans and Other SwitchingBased Networks for Mixing
A Banyan network is a topology of nodes or switches in which there is a single path connecting each input to each output, and there are no internal circuits (or loops).
If the switches are replaced with mixing functions in which each output of each node 170 depends on all inputs, then the same topology can be used for whitebox cryptography.
A 3^{n}×3^{n }base 3 Banyan network has n×3^{(n−1) }nodes and mixes 3^{n }inputs so that every one of the 3^{n }outputs can depend on every input, given suitable choices of functions. That is, if n=2, then there are 6 nodes and the network is 9×9 as shown in
Note that such networks are invertible. Thus, the encoding performed by such a network can be undone simply by creating a mirror network with mirror elements containing functional inverses.
A base 3 Omega network has the same properties as detailed above for a base 3 Banyan network. However, a Banyan network addresses among nodes by repeatedly subdividing the network by the base (e.g., if there are 9 nodes, it first selects among the first third, second third, or third of the nodes, and then selects a third of the three nodes: one specific node).
An Omega network addresses differently. Number the input ports (where each node has a number of input ports specified by the base, and a number of output ports specified by the base) with a 2 digit base 3 number. Then each output port is connected to the succeeding input port found by rotating its number left, circularly, one position.
Other than their manner of interconnection, the networks have very similar properties.
The significance of these networks is that, where, for switching purposes, each node (for a base, b) is a b×b crossbar switch, for our purpose, each node is a function (which would usually be bijective) of the form E.
We can start with linear nodefunctions, and delinearize them by the methods previously described in section 3. This is an alternative to the blocking method for representation, but restricts us to a subset of the linear functions which can be so represented.
The big advantage is the reduction in spatial complexity from O(m_{#} ^{2 }n_{#}) for a given blocking factor, to O(m_{#} log n_{#}): a huge savings in space. It also represents the same reduction in computational time, because a lookup must be done on each SB, and the huge reduction in the number of SBs therefore also represents a huge saving in execution time.
10.3 Compact m×n Nonlinear Boolean Functions
As noted above, memory capacity may be severely restricted and inflexible, as in the case of Smart Cards. Thus, techniques which optimise memory usage and still offer proper protection, are highly desirable.
Method I was already described above under section 4.1. Methods II and III proceed as follows:
10.3.1 Method II: Encoding of BitPosition Permutation Functions Composed into a BitPosition Permutation Network
We can take a series of elements which take in n bits and output n bits, where each output bit is equal to one of the input bits: a bitposition permutation, such as a Pbox in DES.
We can perform arbitrary permutations of bitpositions using elements such as 4×4, 3×3, or 2×2. However, encodings of 2×2 functions are all linear, so let us restrict our attention to those which are at least 3×3.
We can then encode the inputs and/or outputs so that the resulting functions are nonlinear.
When an entropic repositioning is performed in this way, and can be expressed, before encoding, in this form, then if we use switching elements each of which permutes some relatively small number of bits (but 3 or more) bounded above, the space and time complexity of the encoded network is on the order of (n log n). This is much smaller than the quadratic order of the implementation of the same kind of function using Method I.
10.3.2 Method III: Encoding of Small Linear Functions Composed into a Network with Arbitrary Acyclic Connectivity
This is a variant of the above, which includes all of the capabilities of Method II but increases the ability to handle compressing functions (m×n functions in which m is greater than n) and mixing functions where each output bit is to depend on many input bits.
In this variant, instead of small bitpositionpermuting functions, we use small linear operations, possibly compressing ones (m×n where n<m) and we encode their inputs and outputs (thus making them nonlinear). For example, if we connect them into a Banyan network of the appropriate base, we still have an n×n function, but we can make every output dependent on every input in a network of space and time complexity of order (n log n).
The more we increase the depth of such a network, the greater the entropy added and the less restricted are the functions we can compute. Since we can encode an XOR in this fashion, we can therefore compute an arbitrary Boolean function using such a network. (XOR is a “complete” Boolean operator: any Boolean function can be computed by a direct acyclic graph containing only XOR nodes.) However, the size of network required (e.g., as prime implicants) varies on the order of (2^{m }n).
Thus, in the worst case, there is no improvement in representational compactness compared to the full bittabulated representation, and on the contrary, there is an increase in overhead.
10.4 Mixing Entropy among Nodes in a Network of Boolean Functions: Domain Partitions Induced by Functions
If we consider a function from m to n, where n≧2 and m>n, we have:
F: D→R where:
Let us call P the “domain partition of D induced by F”.
If we are given D, R, and P, then there are exactly n! functions F such that F: D→R and the domain partition of D induced by F is P.
Now, it is plain that for any two such functions, say F and F′, there exists a bijection:
B: R→R, such that:
F′=B∘F and
F=B^{−1}∘F′
Interesting and potentially useful situations arise where we can choose a B and F such that F is linear or can easily be constructed by simple nonlinear functions on linear functions. This is because we know how to break a linear function down into a network which we can then nonlinearize (see Method I in section 4.1 above, for example).
The use of this kind of breakdown of a function is: given a reference implementation, and a domain D which is of searchable size, we have some hope of duplicating the reference implementation, but in a fashion which distributes the entropy and functionality quite differently from the reference implementation.
For example, if we consider any 4 bits produced as an output of the Pbox permutation in DES, it is the function of somewhere around 24 bits of input given a specific key and round. If we can implement this function by breakdown into a simple nonlinear function of linear components followed by a bijection, we can break it down into a network in which multiple partial Sbox lookups have been smeared together. A particular way of doing this for Sboxes is presented in section 10.5.1.
Note that the above transform uses m>n. This is done because it is advantageous to change the number of input bits and output bits in a known construction, such as a DES Sbox. If we do not, then irrespective of what bijective encodings we apply to the inputs and outputs, the attacker at least knows which sets of inputs produce exactly the same output. This can reduce the number of possibilities the attacker must consider, thereby speeding up the cracking process.
10.5 Moving Entropy with Full Data Dependency
10.5.1 The ‘Red Path’ Method
We often encounter the following subproblem of the larger problem of cloaking cryptographic and other functions:
Plainly, any approach to the above problem cannot simply narrowly move the information in small bundles, individually switched—it would then be possible to tell exactly where the information moved, which is giving away too much information. Ideally, we would like every output of the switching network to depend on every input of the switching network (where the switching network's job is to reconfigure the information according to a needed switching function).
What we want is a switching network with the following properties:
This results in a big switching network which sends lots of bits in many directions. Some of the inputs and outputs are relevant to the underlying computation, and some are there for the purpose of obfuscation. The paths from relevant input bits to relevant output bits are referred to herein as “red paths”. All of the other paths are present, but irrelevant. That is, the nonred paths are the ones which are present for obfuscation and the red paths are the ones we actually need to do useful work.
The proposal on how to do this is presented briefly in the flow chart of
It would not be necessary to exactly use Banyans, but we do need full connectivity, probably multiple times.
Alternatively, one could lay down simple functions in the switch nodes (e.g., bitwise XOR). This would allow switching and computing to be encoded together.
10.5.2 The Big Bundle Proposal
Another approach to the protection of a standard DES algorithm is presented in the flow chart of
First, at step 200, the initial Sboxes 50 are convert from their 6 bit input, 4 bit output form, to 6×6, each modified Sbox simply containing a permutation of 0 . . . 63.
Next, at step 202, the initial Pbox 52 is converted to accommodate the 6×8 input bits from the Sboxes modified at step 200, and to provide an output 32×6=192 bits wide in 32 bundles of 6. In other words, the full 6 outputs of an Sbox are used wherever one bit of the Sbox would be required.
At step 204, the 36bit Sbox input in the 2nd round is rendered down to k bits, where k≧6. This step must be done such that any imbalances in the round1 Sboxes are overcome; that is, trying to make the buckets equiprobable. This step is then repeated for each subsequent round of the DES algorithm, per step 206.
This technique does not hide the identity of the Sboxes 50, but only hides the key. If we can preserve the secret of the interior coding, then this technique may be effective.
However, if k>6, then a bucketing attack can identify the input values representing the same value. To prevent attack on the bucketing of the outputs of the Sboxes, we must make the implementation balanced. That is, we cannot divide codings into Sbox output buckets evenly. We should arrange the buckets so that the imbalance of the Sboxes cannot be used to mount an attack.
This should also be done with consideration for avoiding the leakage of information due to the relationship among the Sboxes at each round, or among rounds.
11.0 The Information Reconfiguration Problem
An important recurring problem in implementing m×n Boolean functions by networks is the information reconfiguration problem, which is the problem of how to reconfigure the distribution of entropy between the inputs and outputs of a function, or its realization by a network, while preserving an acceptable level of security in the data encodings employed.
11.1 The DES PBox Example
Exemplary instances of the information reconfiguration problem are the implementation of the Expansion permutation QPMd 46 or Pbox permutation 52 in DES.
If we partially evaluate the Sboxes 50 with respect to the key, then a DES round may be presented per the block diagram of
It is fairly easy to protect the functions represented by the Sbox substitution step 210 and the XOR step 212.
For example, for the XOR 212, if we use 8×4 substitution boxes, where four of the eight inputs come from the left side and four from the right, so that we have, in effect, two 4bit inputs and one 4bit output, and if we encode the 4bit inputs and the output independently using bijective mappings, then there are (16!)^{3 }distinct implementations of an encoded XOR function, the vast majority of which are nonlinear. It seems unlikely that an efficient attack can be mounted at the XOR substitution boxes.
Then if, for the Sboxes 210, we use 6×6 boxes, each providing a bijection, where there is a 6×6 bijective encoding of the input, and a 6×4 decoding for the output, then each Sbox 210 is unconditionally secure by itself, since each Sbox table is a permutation of the numbers 0 . . . 63 and by manipulating either the input or output encodings we can produce any arbitrary permutation in the table.
This leaves the problem of modifying the expansion permutation 214 and the Pbox permutation 216 so that the values actually traversing the transformed network do not reveal the underlying values which they represent. This is a more complex problem and is addressed in section 11.2.
11.2 Sample Application: DES
Now, let us consider a specific DES example. Referring to the flow chart of
First, we take all eight DES Sboxes 50 at step 220 and replicate each Sbox 50 sixteen times so that we have a complete set of eight Sboxes for each round of DES. Then for each round, we partially evaluate at step 222, so that the key information is eliminated by changing the contents of the Sboxes 50.
We make no attempt to disguise the inputs of the 1st round of Sboxes 50. However, we encode the outputs as 6bit values, rather than the 4bit outputs of the original Sboxes 50. We can do this so that the contents of each round 1 Sbox contains a table which is simply some permutation of the numbers from 0 to 63 (because 2^{6}=64). Hence, the Sboxes 50 are “unconditionally secure” with respect to the key bits folded into them. Note that the output bits contain all of the entropy of the input bits.
To make this happen, we preferably encode Sboxes 50 at step 224 as follows: we first create a 6×6 Sbox by making the loworder 4 bits of output the same as the unencoded outputs, and prepending the two row selection bits. Since each row is a permutation of the integers from 0 to 15, and hence, encodes a 4×4 bitvector bijection, this guarantees that no entropy is lost. We then encode the resulting 6 bits arbitrarily using an arbitrary nonlinear permutation of 0 . . . 63, and fill the table accordingly.
To get to the next round, we need to do the following: to model the Pbox permutation 52 followed by the bitwise XOR 56 with the left side, and emitting the new left side (a recoded version of the incoming right side and the new right side—the output of the bitwise XOR 56).
We note that the function taking us from one round to the next in the original implementation is an 80×80 bitvector function. If we decode the 6×6 Sbox outputs before input to this function, and code them again for the outputs, then what remains between the decode and the encode is a linear 80×80 function, since the XOR 56 is plainly linear, the Pbox 52 is linear, and the composition of linear functions, or a concat combination of linear functions, are also linear.
A bitwise XOR of two nbit vectors is computed by a matrix of n rows and 2n columns where we multiply the matrix by the concatenation of the two input vectors, and the result is the output vector. The k'th row of the matrix has (k−1) binary 0s followed by a 1, followed by n binary 0s followed by a 1, followed by (n−k−1) binary 0s. In general, for permutations of the inputs and outputs, the matrix would have linearly independent rows in which each row contained only two 1's and each column contained only a single 1.
12. Reconfiguring Substitution Boxes
12.1 Combining SBoxes in DES
Sboxes 50 in DES can be arranged into a ring, in which each Sbox 50 shares two (preexpansion permutation) inputs with its neighbour to the left in the ring, and two (preexpansion permutation) inputs with its neighbour to the right in the ring.
Hence, any pair of neighbouring Sboxes 50 can be combined into a single new Sbox which has ten inputs and eight outputs. The size of such an Sbox is 1K bytes, but we only need 4 of the new Sboxes per round, so 4K bytes provides all of the Sboxes needed for one round.
If we output a mixing linear map of the output, we can still do the XOR and get to the next round via a single 64input Mbox (see
12.1.1 Summary of DES SBox Hiding
The work on hiding SBox lookups and DES in general can be summarized with the following key observations:
We could also:
This could also be done with more than 2 bits out from steps ii) and iii).
After unravelling the permutation:
S1  3  4  5  6  7  8  
has  
in  
puts  
S2  7  8  9  10  11  12  
has  
in  
puts  
S3  11  12  13  14  15  16  
has  
in  
puts  
S12′ has  4  5  6  7  8  9  10  11  
S23′ has  8  9  10  11  12  13  14  15  
S1′  2  3  4  5  6  7  8  9  
has  
S12′  4  5  6  7  8  9  10  11  
has  
S2′  6  7  8  9  10  11  12  13  
has  
S23′  8  9  10  11  12  13  14  15  
has  
Thus, the plan as seen by attacker (using the B) ii) complex variant), is as follows:
The codings used in this system are as follows:
How to execute the entire method:
Another way to add confusion in an implementation involving substitution boxes is to split one substitution box into multiple boxes, such that none of the boxes created by the split has sufficient information to compute the original presplit result, but the information from their combined outputs is sufficient to compute the original presplit result, assuming that we know how their outputs are encoded.
12.2.1 Example: Splitting a DES SBox
Every Sbox, say box B, in the original DES implementation, or the DES implementation after Sboxes have been modified by folding the 6bit portion of the round subkey relevant to that box, has inputs made up as follows:
In the initial rounds, the path of the input bits clearly identifies the Sbox, making it very easy to determine which box is which.
Let us ignore the sharing above and consider splitting one Sbox in isolation, without worrying about shared connectivity.
We have a 6×4 Sbox, S. Let us make an Ubox, U, and a Vbox, V, as follows: U and V are 6×4 substitution boxes. U takes three Sinputs and three inputs not from S. V takes three Sinputs and three inputs not from S. U and V do not share any input bits. We want to be able to determine the Soutput vector from the Uoutput vector and the Voutput vector.
We note that, given the value of the U inputs, there are then 2^{3}=8 possible output values (not necessarily distinct) depending on the value of the three Sinputs which go to V but not U.
Hence, as a first approximation, for a given input vector, U can output:
<R_{0}, R_{1}, . . . , R_{7}>
For V, we can have V output the index i, a number in the range 0 . . . 7, used to index into the 32bit output of U to select a particular 4bit block within it. (This has the bad effect of producing one invariant Voutput, which we do not want because it narrows the attacker's search space.)
Of course we would want to encode all of the inputs and outputs: the above situation would be a preencoding situation. However, even as a preencoding situation, it is not good, because U and V each have three “do not care” inputs: bad news. This kind of situation is highly detectable just by looking at the Ubox and Vbox tables. This gives away too much information narrowing down the attacker's search space.
Let us modify it: we have U output the concatenation of the elements of:
<R_{0}, R_{1}, . . . , R_{7}, E>
that is:
R_{0}∥R_{1}∥ . . . ∥R_{7}∥E
(a 38bit Uoutput vector) where the R's are as above, and E (for entropy) is simply the entire Uinput vector. (We make E be the full Uinput vector because we do not know whether the Rportion of the Uoutput is different for each distinct input vector. Hence, unless we make E the full six input bits, it may lose information.) Now U is nonlossy: there are no do notcare inputs. Moreover, full entropy is preserved across all outputs, even if it would not be preserved by S. Hence, no information can be deduced from the way entropy is lost: it is not lost.
In the same spirit, we have V output i∥j where i is the 3bit value in the range 0 . . . 7 used to index into an R substring of the Uoutput vector, and j is the remaining three bits. As a result, V is likewise entropypreserving.
S1  3  4  5  6  7  8  
has  
in  
puts  
S2  7  8  9  10  11  12  
has  
in  
puts  
S3  11  12  13  14  15  16  
has  
in  
puts  

Thus, we can make an Abox with 8 inputs and 8 outputs, which narrows its left overlapping original box to one of two and the right overlapping original box to one of two possibilities. We know we can represent the Aoutput in 8 bits since only 8 bits went in.
We can also make a Bbox with 8 inputs and 8 outputs, which narrows its left overlapping original box to one of two and the right overlapping original box to one of two possibilities. We know we can represent the Boutput in 8 bits since only 8 bits went in.
We can use overlap, or avoid overlap. If we use overlap, note that we have four bits of overlap—enough to narrow down the alternatives to four.
Alternative overlap is:.

While particular embodiments of the present invention have been shown and described, it is clear that changes and modifications may be made to such embodiments without departing from the true scope and spirit of the invention.
The techniques of the invention have been described with respect to the DES application, but clearly can be applied in many other contexts. The AES (Advanced Encryption Standard) approved under FIPS197 would be a typical example. The linear (or “affine” to use the terms employed in the standard) transformations in AES including the Sbox, ShiftRows and MixColumns transformations could all be obfuscated using the techniques of the invention. The AES is well known in the industry and information is available available. The FIPS standard for AES is available on the Internet at http://csrc.nist.gov/encryption/aes/ and exemplary software code is available at http://csrc.nist.gov/encryption/aes/rijndael/.
It is understood that as decompiling and debugging tools become more and more powerful, the degree to which the techniques of the invention must be applied to ensure tamper protection, will also rise. As well, the concern for system resources may also be reduced over time as the cost and speed of computer execution and memory storage capacity continue to improve.
These improvements will also increase the attacker's ability to overcome the simpler tamperresistance techniques included in the scope of the claims. It is understood, therefore, that the utility of some of the simpler encoding techniques that fall within the scope of the claims, may correspondingly decrease over time. That is, just as in the world of cryptography, increasing keylengths become necessary over time in order to provide a given level of protection, so in the world of the instant invention, increasing complexity of encoding will become necessary to achieve a given level of protection.
The method steps of the invention may be embodied in sets of executable machine code stored in a variety of formats such as object code or source code. Such code is described generically herein as programming code, or a computer program for simplification. Clearly, the executable machine code may be integrated with the code of other programs, implemented as subroutines, by external program calls or by other techniques as known in the art.
The embodiments of the invention may be executed by a computer processor or similar device programmed in the manner of method steps, or may be executed by an electronic system which is provided with means for executing these steps. Similarly, an electronic memory medium may be programmed to execute such method steps. Suitable memory media would include serial access formats such as magnetic tape, or random access formats such as floppy disks, hard drives, computer diskettes, CDRoms, bubble memory, EEPROM, Random Access Memory (RAM), Read Only Memory (ROM) or similar computer software storage media known in the art. Furthermore, electronic signals representing these method steps may also be transmitted via a communication network.
It will be clear to one skilled in these arts that there are many practical embodiments of the DES implementation produced by the instant invention, whether in normal executable machine code, code for a virtual machine, or code for a special purpose interpreter. It would also be possible to directly embed the invention in a netlist for the production of a pure hardware implementation, that is, an ASIC.
It would also be clear to one skilled in the art that this invention need not be limited to the existing scope of computers and computer systems. Credit, debit, bank and Smart Cards could be encoded to apply the invention to their respective applications. An electronic commerce system in a manner of the invention could for example, be applied to parking meters, vending machines, pay telephones, inventory control or rental cars and using magnetic strips or electronic circuits to store the software and passwords. Again, such implementations would be clear to one skilled in the art, and do not take away from the invention.
Cited Patent  Filing date  Publication date  Applicant  Title 

US5892899 *  Jun 13, 1996  Apr 6, 1999  Intel Corporation  Tamper resistant methods and apparatus 
US6668325 *  Jun 9, 1998  Dec 23, 2003  Intertrust Technologies  Obfuscation techniques for enhancing software security 
US7073072 *  Mar 17, 1999  Jul 4, 2006  Axalto Sa  Method to prevent power dissipation attacks on a cryptographic algorithm by implementing a random transformation step 
US7162031 *  Dec 16, 1999  Jan 9, 2007  Nokia Corporation  Method and device for cryptographically processing data 
US20020016918 *  May 14, 2001  Feb 7, 2002  David Tucker  Information security method and system 
EP0981223A2 *  Aug 19, 1999  Feb 23, 2000  Kabushiki Kaisha Toshiba  Encryption/decryption apparatus 
Reference  

1  *  Chow, S. et al. "A WhiteBox DES Implementation for DRM Applications", Oct. 15, 2002. 
2  *  Coron, JeanSebastien et al. "On Boolean and Arithmetic Masking against Differential Power Analysis", 2000. 
3  *  Coron, JeanSebastien. "Resistance Against Differential Power Analysis for Elliptic Curve Cryptosystems", 1999. 
4  *  Daemen, Joan et al. "Bitslice Ciphers and Power Analysis Attacks", 2000. 
5  *  Daemen, Joan et al. "Efficient Block Ciphers for Smartcards", May 1999. 
6  *  Goubin, Louis et al. "DES and Differential Power Analysis, The "Duplication" Method", Aug. 1999. 
Citing Patent  Filing date  Publication date  Applicant  Title 

US7743253 *  Nov 4, 2005  Jun 22, 2010  Microsoft Corporation  Digital signature for network coding 
US7769165 *  Oct 14, 2005  Aug 3, 2010  Microsoft Corporation  Semipublic whitebox cipher 
US8144865 *  Dec 19, 2006  Mar 27, 2012  Stmicroelectronics S.A.  Protection of the execution of a DES algorithm 
US8155306 *  Dec 9, 2004  Apr 10, 2012  Intel Corporation  Method and apparatus for increasing the speed of cryptographic processing 
US8161463 *  Oct 31, 2007  Apr 17, 2012  Irdeto Canada Corporation  System and method of interlocking to protect software—mediated program and device behaviors 
US8171306 *  Nov 5, 2008  May 1, 2012  Microsoft Corporation  Universal secure token for obfuscation and tamper resistance 
US8539596  May 21, 2009  Sep 17, 2013  Cisco Technology Inc.  Security within integrated circuits 
US8666067 *  Nov 1, 2011  Mar 4, 2014  Stmicroelectronics (Rousset) Sas  Cryptographic countermeasure method by deriving a secret data 
US8752032 *  Feb 23, 2007  Jun 10, 2014  Irdeto Canada Corporation  System and method of interlocking to protect softwaremediated program and device behaviours 
US8913745  Aug 5, 2013  Dec 16, 2014  Cisco Technology Inc.  Security within integrated circuits 
US8976960  Apr 2, 2012  Mar 10, 2015  Apple Inc.  Methods and apparatus for correlation protected processing of cryptographic operations 
US9213826  Nov 12, 2010  Dec 15, 2015  Irdeto B.V.  System and method to protect Java bytecode code against static and dynamic attacks within hostile execution environments 
US9286474 *  Apr 11, 2014  Mar 15, 2016  Electronics And Telecommunications Research Institute  Encoding apparatus and method 
US9311489  Mar 7, 2013  Apr 12, 2016  Microsoft Technology Licensing, Llc  Application forensics 
US9363073  Nov 26, 2013  Jun 7, 2016  Stmicroelectronics (Rousset) Sas  Cryptographic countermeasure method by deriving a secret data 
US9503259  Feb 9, 2012  Nov 22, 2016  Irdeto B.V.  System and method for generating and protecting cryptographic keys 
US9509495  Aug 8, 2014  Nov 29, 2016  Samsung Electronics Co., Ltd  Data protection method and apparatus 
US9588756 *  Mar 22, 2012  Mar 7, 2017  Irdeto B.V.  Updating software components in a program 
US9589115 *  Jan 17, 2008  Mar 7, 2017  Panasonic Intellectual Property Management Co., Ltd.  Obfuscation assisting apparatus 
US20060126843 *  Dec 9, 2004  Jun 15, 2006  Brickell Ernie F  Method and apparatus for increasing the speed of cryptographic processing 
US20070022288 *  Jul 5, 2006  Jan 25, 2007  Stmicroelectronics S.A.  Checking of a digital quantity stored in a memory area 
US20070086586 *  Oct 14, 2005  Apr 19, 2007  Microsoft Corporation  Semipublic whitebox cipher 
US20070118746 *  Nov 4, 2005  May 24, 2007  Microsoft Corporation  Digital signature for network coding 
US20070263859 *  Dec 19, 2006  Nov 15, 2007  Stmicroelectronics S.A.  Protection of the execution of a DES algorithm 
US20080208560 *  Oct 31, 2007  Aug 28, 2008  Harold Joseph Johnson  System and method of interlocking to protect software  mediated program and device behaviors 
US20080216051 *  Feb 23, 2007  Sep 4, 2008  Harold Joseph Johnson  System and method of interlocking to protect softwaremediated program and device behaviours 
US20080301448 *  Jun 1, 2007  Dec 4, 2008  Microsoft Corporation  Security Against Corruption for Networked Storage 
US20100115260 *  Nov 5, 2008  May 6, 2010  Microsoft Corporation  Universal secure token for obfuscation and tamper resistance 
US20100180346 *  Jan 17, 2008  Jul 15, 2010  Nicolson Kenneth Alexander  Obfuscation assisting aparatus 
US20120106732 *  Nov 1, 2011  May 3, 2012  Stmicroelectronics (Rousset) Sas  Cryptographic countermeasure method by deriving a secret data 
US20120189119 *  Apr 5, 2012  Jul 26, 2012  Brickell Ernie F  Method and Apparatus for Increasing the Speed of Cryptographic Processing 
US20150074803 *  Apr 30, 2014  Mar 12, 2015  Irdeto Canada Corportation  System and method of interlocking to protect softwaremediated program and device behaviours 
US20150113518 *  Mar 22, 2012  Apr 23, 2015  Irdeto Canada Corporation  Updating software components 
US20150127953 *  Apr 11, 2014  May 7, 2015  Electronics And Telecommunications Research Institute  Encoding apparatus and method 
US20150161363 *  May 17, 2013  Jun 11, 2015  Koninklijke Philips N.V.  Method, system and device for protection against reverse engineering and/or tampering with programs 
CN103081398A *  Mar 31, 2010  May 1, 2013  埃德图加拿大公司  System and method for protecting cryptographic assets from a whitebox attack 
CN103081398B *  Mar 31, 2010  Apr 6, 2016  爱迪德技术有限公司  用于保护密码资产免受白盒攻击的系统和方法 
EP2458774A1  Nov 24, 2010  May 30, 2012  Nagravision S.A.  A method of processing a cryptographic function in obfuscated form 
WO2011057393A1 *  Nov 12, 2010  May 19, 2011  Irdeto Canada Corporation  System and method to protect java bytecode code against static and dynamic attacks within hostile execution environments 
WO2013116916A1 *  Feb 9, 2012  Aug 15, 2013  Irdeto Canada Corporation  System and method for generating and protecting cryptographic keys 
WO2015149826A1 *  Mar 31, 2014  Oct 8, 2015  Irdeto B.V.  Protecting an item of software 
WO2015149827A1  Mar 31, 2014  Oct 8, 2015  Irdeto B.V.  Obfuscated performance of a predetermined function 
WO2016155779A1  Mar 30, 2015  Oct 6, 2016  Irdeto B.V.  Data protection 
WO2016156513A1  Mar 31, 2016  Oct 6, 2016  Irdeto B.V.  Online advertisements 
U.S. Classification  380/28, 713/190, 713/194 
International Classification  G06F21/14, H04L9/28, H04L9/00, G06F11/30, H04L9/32, H04K1/00 
Cooperative Classification  H04L2209/24, G09C1/00, H04L2209/16, H04L9/0631, H04L9/003, H04L9/0625, G06F21/14, H04L2209/08, G06F21/125 
European Classification  G06F21/14, H04L9/06 
Date  Code  Event  Description 

Aug 12, 2004  AS  Assignment  Owner name: CLOAKWARE CORPORATION, CANADA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:JOHNSON, HAROLD J.;CHOW, STANLEY T.;EISEN, PHILIP A.;REEL/FRAME:015036/0243 Effective date: 20040625 
Feb 20, 2012  REMI  Maintenance fee reminder mailed  
Jul 8, 2012  LAPS  Lapse for failure to pay maintenance fees  
Aug 28, 2012  FP  Expired due to failure to pay maintenance fee  Effective date: 20120708 